CODEWITHSUNDEEP
python-3.xtensorflow
Anusha Mehta
Anusha Mehta

Reputation: 99

How to visualize image from the tensor of shape (?,1152,8)?

I am trying to visualize capsule network layers. Following are layers:

conv_layer1=tflearn.layers.conv.conv_2d(input_layer, nb_filter=256, filter_size=9, strides=[1,1,1,1],
                                    padding='same', activation='relu', regularizer="L2", name='conv_layer_1')
conv_layer2=tflearn.layers.conv.conv_2d(conv_layer1, nb_filter=256, filter_size=9, strides=[1,2,2,1],
                                    padding='same', activation='relu', regularizer="L2", name='conv_layer_2')
conv_layer3=tf.reshape(conv_layer2,[-1,1152,8], name='conv_layer3')

the shape of each layer is as follows:

layer_1: (?, 50, 50, 256)
layer_2: (?, 25, 25, 256)
layer_3: (?, 1152, 8)

Here, I can visualize first two layers with random training image. The code for visualization is as follows:

image = X_train[1]
test = tf.Session()
init = tf.global_variables_initializer()

test.run(init) #(tf.global_variables_initializer())
filteredImage = test.run(conv_layer3, feed_dict{x:image.reshape(1,50,50,3)})
for i in range(64):
    plt.imshow(filteredImage[:,:,:,i].reshape(-1,25))
    plt.title('filter{}'.format(i))
    plt.show()

Here, for visualizing third layer, I got the following error:

InvalidArgumentError: Input to reshape is a tensor with 160000 values, but the requested shape requires a multiple of 9216
 [[node conv_layer3_9 (defined at <ipython-input-36-fd98b9e18bda>:20)  = Reshape[T=DT_FLOAT, Tshape=DT_INT32, _device="/job:localhost/replica:0/task:0/device:CPU:0"](conv_layer_2_11/Relu, conv_layer3_9/shape)]]

How to overcome this and visualize layer 3?

Upvotes: 0

Views: 326

Answers (1)

D_negn
D_negn

Reputation: 378

The problem is on the line where you define your third layer conv_layer3=tf.reshape(conv_layer2,[-1,1152,8], name='conv_layer3') The input to this layer conv_layer2 has a shape of ?,25x25x256 which give the 160000 value on the error and you want to reshape that to a shape of ?, 1152x8 which gives the 9216. So that the reshape works the first one should be multiple of the second.

Upvotes: 1

Related Questions